The Computing Community Consortium (CCC) is pleased to announce the third webinar in the Community Chat series: Powering the Solution: Balancing Computing’s Energy Demands with Its Contributions to Sustainability. Join the authors of the recent

Guest Speakers Share Their Expertise

In this one-hour Zoom webinar and Q&A moderated by article author Michela Taufer (University of Tennessee Knoxville), hear key findings from the article alongside three short talks centered around different aspects of the sustainability conversation.

• Chandra Krintz, Professor of Computer Science at the University of California, Santa Barbara, will open with a look at agriculture and ecology, and what balancing computing’s demands against broader societal needs looks like in applied environmental research.
• Ana Radovanovic, Senior Staff Research Scientist and Technical Lead for Energy Analytics and Carbon Aware Computing at Google, will follow with carbon-aware scheduling and how industry is thinking through the same balance. She previously led the team behind Google’s Carbon-Intelligent Computing System, which shifts flexible computing workloads to times and places where cleaner power is available.
• Adam Wierman, Professor of Computing and Mathematical Sciences at the California Institute of Technology, will close with a talk on the community impact of data centers, and whether they are a burden or an asset to the communities that host them.

All speakers will then be available for questions from the audience during a Q&A, joined by the other authors of the article:

• Michela Taufer, University of Tennessee Knoxville
• Elizabeth Bradley, University of Colorado Boulder
• Tevfik Kosar, University at Buffalo
• Siddharth Parida, Embry-Riddle Aeronautical University
• Masood Parvania, University of Utah
• Eve Schooler, University of Oxford
• Bronis R. de Supinski, Lawrence Livermore National Laboratory

Join Us

Date: Thursday, October 8, 2026

Time: 3:00 PM ET

The question around computing’s energy usage has become a global one. This is your chance to hear directly from the experts and make your voice heard. Use the link below to register for the webinar, and feel free to submit questions ahead of time. 

Can’t make it? Register anyway and CCC will send you the recording afterward. We hope to see you there!

Register Here

Tune in to the CCC LinkedIn Showcase Page for updates and more Community Chats like this. Stay connected with CCC for the latest insights, publications, and opportunities to engage by subscribing here.

This material is based upon work supported by the U.S. National Science Foundation (NSF) under Award Nos. 2300842 and 2619366. These awards support the Computing Community Consortium (CCC), a programmatic committee of the Computing Research Association (CRA). Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.

Action is urgent to realize the full potential of non-military drones. The report establishes milestones for 2028, laying the groundwork for its 2035 vision now.

See Key Recommendations for the Future of Drone Computing

Drones as a National Infrastructure

Computing on the Fly envisions a decade in which drones move goods, medical supplies, and information at a scale comparable to national infrastructure investments like highways and the electric grid. Natural disaster detection drones could identify wildfire sources within minutes, medical supply chains could bypass ground congestion to reach rural hospitals, and nationwide fleets could continuously inspect bridges and power lines.
Standing between today’s capabilities and that future is what the report authors call a “capability gap,” where hardware and aspirations are outpacing the software and systems needed to operate safely at scale. Drone research also remains fragmented across communities, which has left the field without a unified vision and put national leadership in the drone space at risk. Closing the gap requires coordination as much as it needs technological advances, as no single sector can close it alone.

Recommendations for Four Communities

The brief organizes its guidance around four groups, each with a distinct role in closing the capability gap:

• Scientific funding agencies, on the shared research infrastructure and program structures the field needs
• Policymakers, on standards, certification, and regulation as technical problems rather than administrative ones
• Industry, on interoperability, security, and what it takes to operate reliably at fleet scale
• Academia, on workforce development and the cross-disciplinary work the field currently lacks

The report identifies significantly scaling domestic drone production as a central, and time-sensitive, goal for this vision. Many of the brief’s cross-cutting recommendations thus revolve around this aim and the different roles each group plays in allowing that to happen.

Read the Key Recommendations Brief

The findings in the full report, as well as this new brief, come from CCC’s 2025 Computing on the Fly: Navigating a Vision for the Future of Drone Computing workshop. The workshop — co-sponsored by the IEEE Computer Society — took place in Washington, DC in December 2025, convening experts across academia, industry, and government to envision the future of drone computing. These conversations were synthesized in the full Computing on the Fly report by lead authors Kevin Butler (University of Florida), Christopher Stewart (The Ohio State University), Nils Aschenbruck (Osnabrück University), Weisong Shi (University of Delaware), Ufuk Topcu (University of Texas), and Deborah Silver (Rutgers University).

To read the most critical recommendations from the report, read the brief below.

Read the Brief

Tune in to the CCC LinkedIn Showcase Page for updates and more resources like this. Stay connected with CCC for the latest insights, publications, and opportunities to engage by subscribing here.

This material is based upon work supported by the U.S. National Science Foundation (NSF) under Award Nos. 2300842 and 2619366. These awards support the Computing Community Consortium (CCC), a programmatic committee of the Computing Research Association (CRA). Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.

The role of the software engineer is fundamentally changing. AI-assisted software development has, in many cases, become a new norm, fostering a demand for unprecedented speed in production. Yet the transition from traditional coding to AI-powered development is not as simple as receiving a boost to productivity. It requires both brand-new skills on the part of the developer and a more fundamental rethinking of how software — especially at scale — is generated. In this rapidly evolving field, new understandings of how to capture user intent, ensure system quality at AI-generative scales, and train the next generation of developers are essential.

Beyond Code Report

Beyond Code: The Workshop

To better understand and address these challenges, the Computing Community Consortium (CCC) held its Beyond Code: Engineering Trustworthy Software Systems with AI at Scale workshop — co-sponsored by the IEEE Computer Society — from February 25-26, 2026 in San Francisco, CA. The workshop brought together experts in artificial intelligence, software engineering, programming languages, cybersecurity, and systems engineering to assess the evolving role of AI in developing complex and large-scale software systems. 

CCC’s newest report contains findings and recommendations from the Beyond Code workshop. The report is authored by Randal Burns (Johns Hopkins University), Sebastian Elbaum (University of Virginia), Gabrielle Allen (University of Wyoming), Nils Aschenbruck (Osnabrück University), Terry Benzel (University of Southern California), William Gropp (University of Illinois Urbana-Champaign), Rick Kazman (University of Hawaii), and Manish Parashar (University of Utah). It presents critical challenges and recommendations for the future of large-scale AI-generated software development.

Central Topics

Participants of the Beyond Code workshop worked to identify current gaps and struggles in the current capacity to generate large-scale AI-generated software systems. In addition to those findings, they collaborated to find the best ways forward given the rapid changes taking place in the area, setting out concrete recommendations for the future of the field. Some of the topics central to the report include:

• The intent gap: natural language as a lossy interface

Natural language is the primary interface to code models, but its ambiguity and lack of formal semantics lose information in translation. 

• Machine representations beyond human-centric syntax

First-generation models are constrained by syntax designed for human readability, which forces inefficient trial-and-error generation loops.

• Quality attributes and formal assurance

Models produce functional logic but fail on performance, security, and numerical accuracy. Generation velocity and volume outpace human auditing capacity and make traditional quality metrics obsolete. 

• Security as an intrinsic property

Goal-driven security interfaces and secure-by-construction generation, rather than security applied after the fact.

• Multi-agent orchestration and deployment

Operating teams of agents complicates things. It introduces coordination friction, composability failures across siloed outputs, and integration liabilities at deployment. 

• Design rationale, maintainability, and legacy systems

As developers distance themselves from code generation, the rationale behind design choices is lost. Codebases become confusing and the capacity to maintain legacy systems diminishes.

• The developer role, deskilling, and education

Generative tools amplify existing productivity differences. They can enhance expert capability while also creating intellectual dependency and deskilling among the less-experienced. The role of software engineer itself is changing to be a multi-agent orchestrator. Curricula and training must follow.

• The research ecosystem

With industry commanding frontier models and compute infrastructure, academia risks relegation to post-hoc evaluation of commercial artifacts.

• Training data transparency and IP governance

Frameworks for data transparency and IP governance for AI-generated code, including membership inference and provenance tracking.

The report poses recommendations for tackling these pressing challenges within the computing research community.

Read the Full Report

The Beyond Code: Engineering Trustworthy Software Systems with AI at Scale report is available now. Its recommendations aim to create an ecosystem where large-scale AI-generated software is complex, interdependent, and, above all, trustworthy. We encourage all members of the computing community to read the full report here.

Tune in to the CCC LinkedIn Showcase Page for updates and more reports like this. Stay connected with CCC for the latest insights, publications, and opportunities to engage by subscribing here.

This material is based upon work supported by the U.S. National Science Foundation (NSF) under Award Nos. 1734706 and 2300842. This award supports the Computing Community Consortium (CCC), a programmatic committee of the Computing Research Association (CRA). Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.

To that end, the Computing Community Consortium’s (CCC) newest report, Computing on the Fly: Navigating a Vision for the Future of Drone Computing, lays out recommendations for maximizing the potential of drone computing. The findings are informed by the December 2025 Computing on the Fly workshop, which convened experts across computing research, industry, and government to craft an informed and realistic, yet visionary, roadmap for ensuring future U.S. leadership in this competitive field.

Recommendations for Funders of Scientific Research

Near-peer competitors are out-investing the U.S. in autonomous systems’ integration, threatening American leadership in a foundational technology advancement. Cross-agency coordination is essential to maintain American leadership in this area. 

In practice, this coordination looks like the National Science Foundation (NSF) leading foundational research while mission agencies like the U.S. Department of Agriculture (USDA), Department of Transportation (DoT), Federal Aviation Administration (FAA), Federal Communications Commission (FCC), and Department of Defense (DoD) ensure applications address real-world constraints and national priorities. 

Equally critical is establishing a national testbed consortium enabling reproducible research across diverse platforms and environments, ideally serving 20 universities and 100+ industry partners. And further, it is necessary to specifically fund integration platforms that compose AI agents into safety-critical systems.

Recommendations for Policymakers

For non-military drones to realize their full potential, policy frameworks must be proactive rather than reactive to the rapidly evolving technology. It is essential to adopt these reactive frameworks across operator licensing and training, vehicle certification and maintenance, and traffic management, spectrum allocation, and airspace design. 

At the same time, regulatory sandboxes are important for establishing safe experimentation frameworks. It is only through such experimentation that drones can enable solutions addressing issues like the $150B annual weather disaster burden while maintaining public safety. These regulatory sandboxes can accelerate innovation while managing risk. 

Most fundamentally, however, regulatory environments must be conducive to drone production and operation at scale. Initial stakeholder sessions detailing the current bottlenecks and barriers could inform critical decisions in the coming years.

Finally, international coordination is key to ensuring U.S. leadership in this strategic domain. Policy harmonization with allies prevents market fragmentation while maintaining security advantages over adversaries. 

Read the Full Report

The full report, including in-depth explanations and rationale for these recommendations, is available below. We highly recommend reading it, and subscribing below for CCC’s upcoming Key Recommendations Brief on this topic.

Read the Report Here

Tune in to the CCC LinkedIn Showcase Page for updates and more reports like this. Stay connected with CCC for the latest insights, publications, and opportunities to engage by subscribing here.

This material is based upon work supported by the U.S. National Science Foundation (NSF) under Award Nos. 1734706 and 2300842. This award supports the Computing Community Consortium (CCC), a programmatic committee of the Computing Research Association (CRA). Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.

The complete list of authors includes: 

• Michela Taufer, University of Tennessee Knoxville
• Chandra Krintz, UC Santa Barbara
• Elizabeth Bradley, University of Colorado Boulder
• Michael Crowley, University of Southern California
• Tevfik Kosar, University of Buffalo
• Siddharth Shiladitya Parida, Embry-Riddle Aeronautical University
• Masood Parvania, University of Utah
• Eve Schooler, University of Oxford
• Utkarsh Shah, Google
• Bronis R. De Supinski, Lawrence Livermore National Laboratory
• Adam Wierman, California Institute of Technology

See the registration at the bottom of this page to hear directly from these authors in an upcoming CCC Community Chat webinar and Q&A.

CCC Task Force Origins

CCC task forces are groups of computing researchers, composed of CCC Council Members and outside experts, who have identified a long-term computing challenge and aim to produce a resource to help guide the community. In 2023, when Michela Taufer and Adam Wierman joined the Computing Community Consortium (CCC) Council, the first discussions of a piece like “Balancing Sustainable Computing and Computing for Sustainability” were already taking place. At the time, Elizabeth Bradley and Chandra Krintz were part of a CCC task force on sustainability in computing; with the additions of Michela and Adam, they worked together to define the scope of the work. 

Partly, this meant reconciling the fact that the very same computing advances that often support sustainability work and research also impact the environment through their significant energy demands. They knew that they wanted that tension to be front and center in their work, and that they wanted the input and participation of members of the broader computing community on this topic. 

To that end, they ultimately invited other members of the community to join them in speaking at two different CCC-funded panels: one at the International Conference for High Performance Computing, Networking, Storage, and Analysis (SC24) in Atlanta, GA in 2024, and the second at the American Association for the Advancement of Science (AAAS) 2025 Annual Meeting in Boston, MA. They supplemented the discussion with panelists from different research backgrounds, as well as connected directly with two different audiences key to the conversation: the developers of high-energy computing technologies and advances, as well as their users.

In Michela Taufer’s words, “We had these visions and these ideas, but we had to find a way to come together and coordinate so that what we were doing became a roadmap for the whole community. And it was CCC that enabled us to come together to do that.”

Balancing Sustainability

In the end, the authors draw an important distinction between sustainable computing and computing for sustainability in the CACM article. Sustainable computing involves minimizing the environmental footprint of digital infrastructure — the digital infrastructure that, notoriously, has led to the proliferation and impact of widespread data centers. Computing for sustainability, on the other hand, means applying computational tools to solve pressing environmental sustainability challenges. The article argues that both approaches are critical for improving computing’s impact on the world, and that coordinated multidisciplinary action — involving computing research, technology development, and governance — is the key to striking that balance. 

For author Adam Wierman, that balance is not only achievable, but deeply urgent. “The energy demand of AI has already outrun what the grid was built to deliver,” he said. “Coordinating computing with the grid by shifting work in time and place is no longer optional.”

CCC’s mission of convening across all sectors of the computing research community also had a significant impact on the ultimate direction of the article. Chandra Krintz highlighted the necessity of working together across disciplines and sectors to make their vision a reality, and how CCC uniquely positioned them to make those connections: “Computing has a unique dual role in sustainability: we must reduce the energy and resource demands of the technologies we create, while harnessing those same technologies to address some of society’s most pressing challenges. Neither goal can be achieved by one discipline alone. Through CCC, we have been able to bring together perspectives from across computing, energy, science, policy, and industry to begin building a shared path forward.”

Join the Conversation

Want to learn more about balancing sustainability and hear directly from the article authors? Join the upcoming CCC Community Chat! 

Powering the Solution: Balancing Computing’s Energy Demands with its Contributions to Sustainability is a one-hour webinar and Q&A that will take place on Thursday, October 8, 2026 at 3 p.m. ET. This is your chance to better understand how computing can contribute to a more sustainable world and ask the authors, who are experts in the field, any burning questions you may have. Register below!

Register Here

Tune in to the CCC LinkedIn Showcase Page for updates and more reports like this. Stay connected with CCC for the latest insights, publications, and opportunities to engage by subscribing here.

This material is based upon work supported by the U.S. National Science Foundation (NSF) under Award Nos. 1734706 and 2300842. This award supports the Computing Community Consortium (CCC), a programmatic committee of the Computing Research Association (CRA). Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.

The renewal means CCC’s visioning activities with the computing research community will continue, supporting the evolving field through 2030 on this new grant.

CCC’s Mission

What are the next big computing research challenges — the ones that will define the future of the field, galvanize researchers, and spark investment and public support?

CCC mobilizes the computing research community to address these challenges. Since 2006, CCC has served as a bridge between computing experts across disciplines and sectors, convening them for visioning activities where they are asked to not only envision the future of computing, but also the path there. These collaborative efforts between academia, industry, and government — and the reports, white papers, and other resources that result from them — leverage the collective expertise of the field to set long-term computing research trajectories in motion and address emerging national and global challenges. 

What to Expect Over the Next Four Years

• Visioning workshops. These invitation-based, multi-day workshops assemble a group of experts around a specific research frontier. Topics originate from the CCC Council or from the community through CCC’s visioning proposal process. We encourage any member of the computing research community to submit a proposal!
• Reports, briefs, and white papers. CCC reports and briefs bring the findings of those workshops to the community. The recommendations within them inform researchers, industry folks, and policymakers about the focused work and funding required to serve the future goals of the field. CCC white papers do similar work, providing community-based expertise and recommendations without pulling from a fully-fledged workshop effort.
• Responses to federal Requests for Information (RFIs). When federal agencies release RFIs on computing research priorities, the CCC Council, working with CRA Government Affairs, submits a consolidated response. This is one of the most direct paths for community expertise to inform federal policy.
• Community Chats. CCC’s webinar series brings report authors, workshop organizers, and subject matter experts into conversation with a broad audience. These Community Chats span wide ranges of subjects within computing, are open to everyone, and are recorded for later viewing.
• Medium posts. CCC’s presence on Medium brings big computing questions to light. Expect accessible writing on computing findings and emerging topics.
• Blue Sky track sponsorships. CCC sponsors Blue Sky Ideas tracks at conferences across computing research, creating space for visionary and forward-looking papers. Sponsorship includes travel support for winning authors.

Get Involved

Two CCC projects are already on the books under this new award. These include our upcoming Full Stack Sustainability: From Silicon to Software workshop in Austin, TX — co-sponsored by the IEEE Computer Society — as well as an open Community Chat on October 8, 2026 called Powering the Solution: Balancing Computing’s Energy Demands with its Contributions to Sustainability (register here). Additional future visioning topics in the works include agentic AI for hardware design and quantum computing.

CCC strongly encourages members of the computing community to subscribe here and follow the CCC LinkedIn Showcase page to stay in the loop about our latest resources, events, and ways to get involved. For individual expert perspectives and commentary on computing that may not fit in a workshop report, follow CCC on Medium.

CCC looks forward to another four years of visioning with the inspiring computing research community!

Key Challenges in Advancing Drone Computing 

The report identifies twelve key technical challenges encompassing the ways in which drone computing must advance to reach its fullest potential in the next ten years. These areas include:

• Scaling to millions of drones
• AI intelligence and assurance
• Edge-cloud continuum and real-time coordination
• AI autonomy and agentic systems
• Data, training, and validation infrastructure
• Critical infrastructure protection
• Building reliable fleets from non-deterministic agents
• Trust, security, and distributed authentication
• Next-generation drone networks
• Human-AI partnership and scalable insight
• Standards, certification, and regulation
• Workforce development and education

The Role of Industry

The specific ways these challenges can be addressed are then outlined in several computing milestones to be completed by 2035: 

• Create simulation-to-reality pipelines that help bridge the sim-to-real gap 
• Establish integration testbeds (physical + digital) supporting graduated complexity, realistic failure injection, and shared milestones for reproducible research
• Establish scalable and performant drone-based entity tags and key infrastructure for identification and tracking at device level 
• Develop real-time-aware autonomy frameworks enabling analyzable composition and runtime adaptation and allowing drones and edge nodes to negotiate sensing rates, model fidelity, and control frequencies under enforced safety constraints and predictable timing guarantees

Industry plays a critical role in advancing these milestones — and, crucially, in maintaining U.S.  technological leadership in the area of autonomous systems broadly. Below, see the report’s specific recommendations:

• Launch regulatory sandboxes

Precedents in the wireless and automotive sectors show that standards for interoperability could reduce integration costs by 40-60%. They would both speed up the time-to-market as well as allow smaller companies to participate.  Working with policymakers, industry also must commit to end-to-end pipelines for the development, deployment, and low-cost operation of drones. 

• Participate in testbed consortia

Doing so would validate approaches on real hardware and influence standards development.

• Share integration experiences and failure mode learnings

This information would inform research directions and prevent others from repeating costly mistakes. 

• Invest in workforce development

This future requires interdisciplinary systems engineers. Partnering with universities can help make sure graduates have essential skills for industry participation in this emerging area. 

• Expand domestic commercial drone manufacturing and coordinate internationally to preserve U.S. leadership

Few U.S. companies produce commercial drones capable of executing the advanced technologies outlined in the report. Expanding domestic manufacturing capacity would strengthen supply chain resilience, address concerns around sovereignty, and enable rapid iteration and learning that is essential for a competitive advantage. To that end, U.S. industry participation is essential to maintain technological leadership as international competitors aggressively pursue autonomous systems markets, especially in Asia and Europe where regulatory experimentation is advancing rapidly. 

Read the Full Report

The full report, including the in-depth explanation and rationale for these recommendations, is available below. We encourage all members of the computing community to read it.

Read the Report Here

Tune in to the CCC LinkedIn Showcase Page for updates and more reports like this. Stay connected with CCC for the latest insights, publications, and opportunities to engage by subscribing here.

Selected Research Projects

Computational Animal Communication

Led by Kenny Q. Zhu, University of Texas at Arlington

Recent advances in self-supervised learning, multimodal foundation models, and web-scale data collection create an unprecedented opportunity to transform this landscape. The Arlington Computational Linguistics Lab at the University of Texas at Arlington is pioneering a new computational paradigm for decoding animal communication from large-scale, minimally annotated multimodal data. Initial studies of domestic dogs and cats suggest that their vocalizations contain discrete acoustic units and reusable patterns that encode behavioral, social, and environmental context, challenging the traditional view that animal vocalizations are largely holistic signals.

This vision extends far beyond understanding companion animals. It seeks to establish Computational Animal Communication as a new interdisciplinary field that develops general computational theories, algorithms, and foundation models for discovering, representing, and interpreting communication systems across species. Such advances could transform animal welfare, conservation, and human-animal interaction while providing a unique window into the evolution of intelligence and communication. More broadly, uncovering alternative computational principles underlying biological communication may inspire fundamentally new paradigms for artificial intelligence that complement language-centric approaches and expand the foundations of intelligent systems.

Punishing Agents

Led by Maria Gini, University of Minnesota

What do you do when an agent cannot be trusted? Expel it from the group? Put it in a prison? None of the methods human societies use seems to be appropriate for agents.  If an agent is removed, another can be created to replace it. Same if an agent is put in a prison. Unless the method used to create agents guarantees that only collaborative agents will be created and/or will be selected, it does not seem to be possible to guarantee that no agent will be able to infiltrate the group.

Attempts have been made to have conventions for agents that negotiate on tasks to do, but they do not work for all types of tasks. The impact on societies of agents of rewards and punishments has been studied showing that they do not work. Reputation seems to be more useful to foster social behavior.

Will things be different for physical agents (i.e., robots)? It is much easier to control a physical agent, take it out of a group, and replace it with a new agent that has been tested for compliance with the rules. Much more research is needed.

Emerging Technologies: Digital Accessibility

Led by Vlad Roznyatovskiy, Samsung Research America

This research initiative proposes a systems-oriented agenda for accessibility spanning hardware, operating systems, programming tools, artificial intelligence, security, and human–computer interaction. Its central goal is to develop computing systems that can adapt to differences in sensory, motor, cognitive, linguistic, and situational ability without requiring users to conform to a single assumed mode of interaction. Rather than treating accessibility as a downstream interface problem, the initiative will investigate how accessibility requirements can be embedded into system architectures, data models, development frameworks, evaluation methods, and procurement standards from the outset.

A major focus will be the relationship between accessibility and language inclusion. Interfaces centered on written English, conventional keyboards, and standardized commands exclude not only people with disabilities, but also users across languages, literacy levels, dialects, and cultural contexts. Multimodal systems that combine speech, gesture, vision, symbols, translation, personalization, and context-sensitive assistance can reduce both disability-related and linguistic barriers. In this sense, accessibility research offers a pathway toward computing that is less English-dominant, more adaptable, and more globally usable.

The initiative will pursue new interaction techniques, adaptive and multimodal system architectures, accessibility-aware software engineering tools, community-governed datasets, and evaluation frameworks that measure real-world task completion across diverse populations. By positioning accessibility as a driver of foundational computer science research, this program aims to produce systems that are more inclusive, resilient, intuitive, and capable of supporting a wider range of human abilities and forms of communication. The expected outcome is not only broader access to computing, but a new generation of interfaces and systems whose flexibility benefits all users.

Stay in Touch with CCC

CCC looks forward to hearing more of the inspiring work shaping the future of computing. We encourage all members of the computing community to subscribe to CCC here to be notified of more opportunities like this, and to follow our LinkedIn Showcase Page for the latest CCC resources and updates.

Key Challenges in Advancing Drone Computing 

The report identifies twelve key technical challenges in which drone computing must advance in order to support drones as part of our national infrastructure in the next ten years. These areas include:

• Scaling to millions of drones
• AI intelligence and assurance
• Edge-cloud continuum and real-time coordination
• AI autonomy and agentic systems
• Data, training, and validation infrastructure
• Critical infrastructure protection
• Building reliable fleets from non-deterministic agents
• Trust, security, and distributed authentication
• Next-generation drone networks
• Human-AI partnership and scalable insight
• Standards, certification, and regulation
• Workforce development and education

Critical Paths for Researchers

Academia plays a critical role in many of these challenges. The following are key computing milestones identified for the next decade to maximize the impact of non-military drones: 

• Create simulation-to-reality pipelines that help bridge the sim-to-real gap 
• Establish integration testbeds (physical + digital) supporting graduated complexity, realistic failure injection, and shared milestones for reproducible research
• Establish scalable and performant drone-based entity tags and key infrastructure for identification and tracking at device level 
• Develop real-time-aware autonomy frameworks enabling analyzable composition and runtime adaptation and allowing drones and edge nodes to negotiate sensing rates, model fidelity, and control frequencies under enforced safety constraints and predictable timing guarantees

Alongside industry, funding, and policy recommendations to reach these milestones, the report outlines necessary shifts and advancements specifically for researchers to help reach the 2035 vision:

• Break disciplinary silos to address the systems integration talent gap
  • Fewer than a dozen graduate programs currently produce engineers with combined AI-aerospace-security expertise
  • Autonomous drone research demands collaboration between traditionally separated communities in robotics, networking, AI, security, and human-computer interaction 
• Engage with real deployments to understand practical constraints
  • In particular, the labor shortage in aviation and infrastructure inspection creates immediate demand for autonomous solutions 
• Share datasets and frameworks to accelerate collective progress and avoid duplicative effort across institutions

Read the Full Report

The findings of the report are informed by discussions at the CCC Computing on the Fly visioning workshop held December 1-2, 2025 in Washington, D.C. — generously supported by the National Science Foundation (NSF) and IEEE Computer Society — which convened 47 cross-sector experts to articulate a vision for drone development. These discussions resulted in the final Computing on the Fly: Navigating a Vision for the Future of Drone Computing report, authored by Kevin Butler (University of Florida), Christopher Stewart (The Ohio State University), Nils Aschenbruck (Osnabrück University), Weisong Shi (University of Delaware), Ufuk Topcu (University of Texas), and Deborah Silver (Rutgers University).

We encourage all members of the computing research community to read the full report at the link below.

Read the Full Report Here

Tune in to the CCC LinkedIn Showcase Page for updates and more reports like this. Stay connected with CCC for the latest insights, publications, and opportunities to engage by subscribing here.

Last week at the CRA Summit in Prior Lake, MN, members of the CCC Council Executive Committee hosted

The format of the session was built around a card game CCC developed for this kind of brainstorming called Speculative Execution. In small teams, participants drew from two decks: one featuring computing substrates like quantum or biological computing, and one featuring application domains ripe for transformation, from agriculture to email. Each team combined whatever cards they drew and pitched a vision for how that substrate could address a real problem in the application, with the crowd serving as both audience and judge.

Speculative Execution Winners

The session did what it set out to do. It produced out-of-the-box ideas that might not fit neatly into a grant proposal, brought researchers from different fields into direct conversation, and generated a fair amount of fun along the way. But across the two rounds, two particular pitches stood out.

Round one went to Carla Brodley (Northeastern University), Rachel Pottinger (University of British Columbia), Alan Marchiori (Bucknell University), and Christine Julien (Virginia Tech), whose pitch imagined a rapid response temporary robot pothole filler. Their team proposed autonomous units that could deploy to fill potholes on a temporary basis, buying time and improving safety until permanent repairs by humans could be scheduled. The substrate they chose was Robotics, and the application they wanted to improve was Infrastructure Maintenance.

Round two went to Zachary Tatlock (University of Washington), Casey Deccio (Brigham Young University), Stefan Andrei (Cleveland State University), Jen Markoff (University of Washington), and Antony Hosking (The Australian National University), who pitched executable email. Their concept turned email from something to be sorted through into something where only the most relevant messages to you would be sendable. Their substrate of choice was Heterogeneous Computing, with an application to Email.

Both pitches captured what the session was after: unlikely combinations that push researchers to think in long-term, unfamiliar directions.

Download the Game

Want to try the game with your own computing department? The card game, Speculative Execution, is available as a PDF here. The PDF includes instructions for gameplay as well as guidance for teams as they develop the pitches for their ideas.

Tune in to the CCC LinkedIn Showcase Page for updates like this. Stay connected with CCC for the latest insights, publications, and opportunities to engage by subscribing here.